Georgia's ERC Expanding an Unproven Solution

As Georgia's Education Reform Commission nears its deadline for recommendations, we are seeing more specific proposals emerge, including plans to create a new Student Scholarship Organization--one with somewhat greater accountability for academic performance with means testing to assure that funds benefit Georgians with the greatest needs. [A draft of their recommendations is posted here].

The reaction to these proposals has been...predictable:

Jason Bedrick of the CATO Institute calls the ERC proposals "A Solution in Search of a Problem." His blog post assails the ERC recommendations as unnecessary because the current program works just dandy as is. His arguments are worth noting as examples of what passes as support for a program that is essentially without evidence of its effectiveness.

Bedrick asserts, "According to the most recent data, three-quarters of scholarships statewide were awarded to students from families earning less than $62,202 annually."  This statement is not as definitive as it might seem. Looking at what is actually the most recent data, the Calendar Year 2014 Qualified Education Expense Credit Report, the numbers seem to bear out what Bedrick asserts. If one sums all the numbers for quartiles 1, 2, and 3 and divides by the total for all four quartiles, you do arrive at 73.18%.  The problem is 73.8% of what?  Georgia counts the number of scholarships but reports only the number of families receiving scholarships by quartile.

Why does this even matter? Glance at the yellow highlighted cells in this sheet. Each of those cells represents a mismatch between the total number of scholarships awarded and the sum of the number of families presented in each quartile. Georgia GOAL's number match up exactly, but Georgia Student Scholarship Programs's reporting doesn't account for some 495 individual awards--30% of their total.  There is absolutely no public accounting that accurately represents to what extent low-income Georgia students benefit from the current SSO program. As noted by the Andrew Young School of Policy Studies,

Regarding income, SSOs currently report the number of families receiving scholarships, but not the total number or amount of scholarships, by quartile of Georgia adjusted gross income. This reporting provides little detail regarding the income distribution of scholarship recipients and no information on the distribution of scholarship dollars. Georgia’s Tax Credit Scholarship Program (Buschman & Sjoquist, 26)

Bedrick holds up the Georgia GOAL Scholarship Program as an exemplar of how SSOs serve the most needy Georgians without the burden of specific income requirements. Unarguably, GOAL is the most noble implementation of what is a flawed program. But if anything, it is an outlier. Four SSO awarded more than 1000 schlarships for calendar year 2014: Georgia GOAL(4314), A Pay It Forward (2114), Apogee Georgia (1832), and Georgia Student Scholarship (1500). Bearing in mind that the quartile distributions are by family and not scholarship recipient, we still see striking disparities in which income groups benefit from each SSO:

[Click to enlarge]

Georgia GOAL is not representative of how SSOs distribute student scholarships. Also, a review of the 2014 Qualified Education Expense Report reveals some smaller SSO that award 84%, 94%, 100% of their scholarships to families earning over $61,953.00.

Contrary to how they have been presented by some, Georgia SSOs do not consistently provide support to Georgia students requiring the greatest assistance. Current law in Georgia provides that SSO
1021 In awarding scholarships or tuition grants, shall consider financial needs of students
1022 based on all sources, including the federal adjusted gross income from the federal income
1023 tax return most recently filed by the parents or guardians of such students, as adjusted for
1024 family size.
While they might consider income, the current law provides no requirement that SSOs actually relate scholarship awards to lower income students, and as noted by the Andrew Young School of Policy Studies, the reporting requirements provide "no information on the distribution of scholarship dollars."

The tax credit scholarships have been sold as a ticket out of low-performing school for low-income students, but the history of the program creates serious doubt as to whether we have met that expectation. [For that record, check these sources: "Millions of Tax Dollars Diverted to Private Schools," "What a Scam: Poor Kids' Money Snatched by Private Schools!" and "A Failed Experiment: Georgia's Tax Credit Scholarships for Private Schools."]

The Expanding Educational Options and School Choice Subcommittee has also stirred a hornet's nest by having the temerity to suggest that scholarship recipients be tested to measure the quality of the private schools cashing the checks. The subcommittee includes among its recommendations:

Require each SSO to ensure that participating schools that accept its scholarship shall: annually administer either state achievement tests or nationally norm-referenced tests that measure learning gains in math and language arts to all participating students in 3rd, 5th and 8th grades;

Similarly, the Andrw Young School of Policy, notes a dearth of evidence about the current SSO program:
The data that would be necessary, or at least very useful, to conduct an evaluation include:

• The public schools attended by scholarship recipients prior to award of the scholarship. These data would be used to estimate the switching rate and to evaluate the quality of the public schools the students are leaving.
• To explore the improvement in student performance, it would be useful to have scores on comparable standardized tests for the private and previously attended public schools. Private schools are not required to use the same standardized tests that public schools use, but available performance information could be used to judge the benefits in terms of quality of the private schools relative to the public schools the scholarship students had been attending.
• The characteristics of scholarship recipients is an important aspect of any evaluation of the program. Thus, the distribution by race, gender, family income, and age (grade) of the scholarship recipients would be important data to report.
  Georgia’s Tax Credit Scholarship Program (Buschman & Sjoquist, 25).

(I would go further to state that these students shall take the same tests as would their counterparts in the public schools they left to best assess if they are achieving gains in learning.) The focus of education reform in Georgia has been on CCRPI and performance on state assessments as the sole measure of effective schools. The state has defined a successful school as one that produces growth in student learning. Yet the state has enabled millions of dollars in SSO funds--tax credits that reduce state revenue--to flow into private schools without any accounting for whether the students actually benefit. 

Bedrick states that imposing a test on these students would restrict parent's choice: "A testing mandate would force schools and parents that are ideologically opposed to standardized testing to choose between their principles and their pocketbooks." And to that I would have to say, "yep." If state dollars flow to a school, then that school should demonstrate effective practice and instruction through the same measure imposed on public schools.

Louisana provides an excellent case in point: the state requires state assessments for students using state vouchers to attend private schools. And what did Bobby Jindal learn?

The academic performance of Louisiana's voucher students remained low in 2013-14, according to Education Department data released Monday. Of 126 private schools accepting the publicly funded tuition subsidies, 23 posted scores low enough to prevent them from accepting new voucher students next fall.

Less than half the voucher students who took the LEAP and iLEAP exams passed: 44 percent. In public schools, 69 percent of students passed. If the voucher students were a school system, they would be tied for fifth-worst in the state, with Bogalusa and Baker.

The data shed new light on the performance of the Louisiana Scholarship Program. Gov. Bobby Jindal has advocated vouchers as ways to give parents more education choices and to rescue students from failing public schools.
Louisiana voucher students perform poorly in 2013-14; proponents praise gains (Nov 13, 2014).

And in this one respect, Georgia could learn a lesson from Louisana. It is the same thing principals tell new teachers: "If you are going to expect it, you better inspect it." In Georgia, we have to this point chosen not to gather information on the effectiveness of the schools to which these students transfer.

Without a common assessment, the state, SSO donors, and parents do not have an objective evaluation of the quality of the instructional program benefiting from SSO funds. Without lifting the legislated prohibition on gathering and reporting specific information on student race, gender, and family income, we will continue to lack visibility into who actually benefits from this program. What is  most evident about Georgia's tax credit scholarship program is that we know very little about how effective it is or even who benefits from it. Bedrick contends that the new program proposal is "a solution in search of a problem." The new proposal is even worse than that: it is an expansion of a program that has gone far too long without critical evaluation.